Power consumption and compliance with standards
When it comes to power consumption, there are actually no big secrets, because what NVIDIA specifies as TGP is largely complied with. With plenty of 16 watts in idle you are in the range of many board partner cards of the GeForce RTX 2070 Super to RTX 2080 Ti, even if relatively high. The partial load ranges are realized relatively economically, which could also be due to the better controlled clock rate. However, under real full load in the Witcher 3 at Ultra-HD and full warming, the card reaches around 320 watts, which is what is stated in the specs. The following table now shows the maximum values of the respective load scenarios.
Now let’s take a look at what NVIDIA and Palit have given the card in the BIOS in reserve. The 340 watts aren’t much, but they are still something (109% reserve at the power limit).
Let’s now move on to the load on the motherboard slot, whose load is specified by the PCI SIG as 5.5 amps. This results in a maximum power of 66 watts at 12 volts. You can see very clearly that under full load again slightly more than 5 amperes are reached, which is of course permissible. But it is still much more than most Turing cards, probably also to relieve the 12-pin connector of the 12-volt rail.
Transients and power supply recommendation
As I have already proved in detail in my basic article “The battle of graphics card against power supply – power consumption and load peaks demystified”, there are also higher loads in the millisecond range, which can lead to unexplainable shutdowns in the case of unfavorably designed or improperly equipped power supplies. The TBP (Typical Board Power) measured by the graphics card manufacturer or the reviewers alone does not really help in this case for a stable system design.
Peaks with intervals between 1 and 10 ms can lead to shutdowns in very fast reacting protective circuits (OPP, OCP), especially in multi-rail power supplies, although the average power consumption is still within the standard. For this card, I would therefore calculate at least 420 to 450 watts as graphics card load for the normal OC, proportionally to the total secondary power consumption of the system, in order to have enough reserves for the worst case scenario. A short excerpt with higher resolution shows us now the 20 ms measurements (10 μS intervals), as I let them run automatically for value determination:
The power consumption in gaming and torture test in detail
First I compared the power consumption during the Gaming Loop. The power limit of the graphics card is 320 watts
The problem with this resolution of 20 ms is that you can’t see the right spikes either, because they get lost in the averaged acquisition and are simply accumulated broken. Let’s therefore switch to the microsecond range, where NVIDIAS Boost also operates. And lo and behold, the spikes are now also clearly visible. Exactly here we see the real difference with the spikes! While with the old driver it was often between 580 and 600 watts, it is now suddenly at least 70 watts less and also the characteristic and frequency decreases strongly! So we can see very well where the missing 4 watts have gone. Not on average, no, but predominantly in the maximum peaks! This is also one of the reasons why the loss of performance should turn out to be manageably small in the end. Control rage slowed down, visible.
In the stress test, one sees a significantly more relaxed control curve, albeit with various slumps:
Where these dips come from, shows and the analogous to the boost running measurement in the microsecond range, where you can also see the self-protection very nice:
Kommentieren